Our study showed that the introduction of ASCIPC had positive impacts on glycaemic management for patients on subcutaneous insulin therapy during admission. More work is required to reduce the rate of insulin omission and to improve further glycaemic control for inpatients.

The aim of this paper is to examine the prescribing patterns and cost of various formulations of metronidazole in a hospital setting over a 3-month period. Oral metronidazole has high bioavailability (98.9%) with peak plasma concentrations averaged at 2.3 h after dosing. Despite the high bioavailability of oral metronidazole, many patients continue to receive metronidazole intravenously when they are suitable for oral preparation. An audit of 120 consecutive patients prescribed metronidazole was conducted at the Liverpool Hospital, NSW, from March to July 2005. There were 65 men and 55 women (age 18-93). Of the 120 patients, 16 were on oral, 1 on rectal and 103 were on intravenous metronidazole. Treatment was initiated based on clinical diagnoses. Potential pathogens were subsequently identified on only 21 occasions. The use of metronidazole as an oral preparation was contraindicated in 27 patients (22.5%) who were nil-by-mouth. Of these, rectally administered metronidazole was contraindicated in only eight patients. The average course of intravenous metronidazole was 8.0 +/- 9.7 days (mean +/- SD). The total number of intravenous metronidazole treatment days was 824. Oral metronidazole would have been possible in 618 out of the 824 days. The estimated cost to administer each dose of oral, suppository and intravenous forms of metronidazole is $A0.11, $A1.34 and $A6.09 respectively. Thus, substantial savings could be achieved if oral metronidazole were to be administered whenever possible. The early use of oral or rectal metronidazole should be encouraged when there are no clinical contraindications.

reported to be as high as 23%. 3 Crop et al. also reported that hypokalaemia had a frequency of 12% in a general hospital population. 4 At variance with these data, Janko et al. observed that 3.5% of hospitalized patients had hypokalaemia, severe hypokalaemia being reported in only 0.5% of the cases. 5 However, the results of this investigation, which is the largest so far reported in a general population of both outpatients and inpatients, indicate that hypokalaemia is a relatively common finding in inpatients, especially in those hospitalized in surgical and internal medicine wards. However, the frequency of critically low potassium values (<2.8 mmol/L) that would require urgent clinician notification is much lower, approximately 1.4%. We have also shown that the frequency of hypokalaemia and severe hypokalaemia differs dramatically according to the healthcare setting, being the lowest in outpatients.